First Impressions: Bonita Open Solution

Written by Nathaniel Palmer.

Bonita Open Solution represents a complete BPMS framework, with a comprehensive modeling and design studio, portal-oriented user environment, and tunable execution engine.       

Bonita Open Solution (BOS) 5.6 is the latest release of an open source BPMS from BonitaSoft, available in both a free downloadable version and as part of paid ”subscription packs” including access to live support and enhanced product capabilities.

The freely downloadable version of BOS provides the complete framework of the BPMS, consisting of the Bonita Studio, for complete process and initial application design with BPMN 2.0 based modeling, the Bonita User Experience (User XP) design module which follows an email-like worklist metaphor and portal experience, and the Bonita Execution Engine which provides the run-time execution environment. All versions include a library of integration adaptors, as well as the ability to model, build, deploy and manage process-driven applications.

The subscription packs provide value-added capabilities focused on traditional enterprise deployment requirements, such as various features for application performance monitoring and optimization, document generation, user profiles and resources for distributed development, integration ‘wizards’ and enhanced multi-tenant capabilities.

 

The Short History of Open Source BPM

Re-examining open source BPM takes me back over 5 years on my own journey with this space. In 2005, I could no longer ignore the advice of my monthly column in BPM.com’s erstwhile sister publication, Open Source Journal. I left a well-paid, low-stress and by at least some measures prestigious job at a NYSE traded firm to found a start-up named “Open Source 3.0” (incorporated and better known by the less presumptuous “OS30”) that offered none of the aforementioned benefits. Yet I could wait on the sidelines no longer. Open Source 1.0 had already transformed the data center and made millionaires of Linux developers (or at least the venture capitals who backed them.)

“Open Source 2.0” was by this time already mature as a rich ecosystem of OSS alternatives to every conceivable software component and application category, from ERP and CRM to EAI and DBMS. The smart money at that time (the next generation of expectant millionaires, or perhaps billionaires) was betting on service providers who could relieve IT departments and systems integrators from the onerous burden of regression testing the kaleidoscope of ever-changing builds and new releases of open source components by selling subscriptions to pre-integrated, value- added “LAMP Stacks” (Linux, Apache, MySQL and PHP) and other core open source infrastructure. This model may be been the next big thing, and no doubt was needed to get commercial enterprise IT departments on-board with broader open source adoption. This model may have been far too prosaic for my tastes.

Eli Whitney arguably launched the Industrial Revolution, but nobody got excited about interchangeable parts. The excitement came from cheap guns of predictable quality, watches that kept time but didn’t cost a month’s salary, machines that could be repaired rather than replaced when they broke. The innovations of this era – what offered real value creation and stirred the hearts of men – were found in what could be made from the combination of standards and mass-produced (thus commonly available) components, not the components themselves.

OS30 was the anticipation of how process standards and readily available components combined would transform the face of software-enabled business innovation. The leverage for this, as clear to me then as it is today, was to be open source BPM. By this time BPMN and XPDL, while still early generation, had already reached critical mass. All that remained was the business-class open source BPM framework to leverage the growing array of already proven components at the top and bottom of the stack.

Commercially-licensed BPM vendors were already (and in fact still are) heavily reliant on these pieces for the interface, builds, testing, run-time, reporting, integration, rules – everything but the engine which makes it all work. The interchangeable parts were there, but I wanted the guns.

For years I waited. Yet open source BPM remained stuck in the doldrums of either limited functionality versions of proprietary tools (a cruel tease at best) or otherwise open architecture, standards-based tools which were far more homebrew than business class. There were a few promising exceptions, yet none reached critical mass to realize the potential already delivered by commercial vendors. None offered a value proposition better than “cheap” and even that was limited to initial license cost, which was too small a part of the equation when picking the platform on which to bet your business. So like a jilted groom, I waited in vain.

While later than I had hoped for, Open Source 3.0 may have finally arrived. If so, its name is “Bonita.”

The core advantage expected from community-based development is open source software that is not simply cheaper, but specifically that which represents a better product.           

Democratizing BPM: Bonita Open Solution 5.6

In Eric Raymond’s seminal collection of essays packaged as “The Cathedral and the Bazaar” (or “CatB” for you geeks) the core message, the ethos of community-based development, is that open source software is not simply cheaper, it’s better. Linux is better than Solaris, than AIX, or HP-UX. Not at first perhaps, but it is now. Commercial, closed-source UNIX survives only as a vertically-integrated component of a proprietary hardware stack. The ‘bazaar’ produces a better product than the commercial ‘cathedral’ can keep pace with. So why isn’t open source BPM better than commercially developed frameworks? I have my theories, but the point is, in just about every instance it’s not.

What sets apart Bonita from most open source BPM projects is that it is in fact, better. As a blanket assessment, of course, no single product is universally better than the rest. The best qualities of the leading BPM Suites set each apart under different circumstances. Yet the differentiating gaps have shrunk between them in the last few years.

Filling the gaps is largely common functionality that combined defines the concept of a suite – collections of thoroughly-integrated capabilities which require engineering teams and product management to produce. The final differentiating qualities, new areas of functionality and improvements on the existing, requires either substantial in-house resources, or a critical mass of users and developers empowered and incented to provide these back to the community.

The latter is the concept of the “bazaar” and why open source, under the right circumstances, can and should be better than the alternative. Commercial open source is about leveraging the best of both – combining managed product development direction with “crowdsourced” software components.

Two great examples of these are the design-time environment and integration adapters. Although there are great piecemeal open source modeling components, there are no truly enterprise class open source modeling suites. This is something that requires the stewardship of a design team, and this is where much of the BPM competitive landscape has been differentiated in recent years.

 Figure 1: Bonita BPMS Notional Run-Time Architecture

           

On the flip side, the market for integration adapters has largely commoditized in recent years. Maturation of JEE and XML based standards has made the development of adapters much easier, thus the premium available to those developed commercially has shrunk substantially, and consequently few product vendors can justify the cost of developing and maintaining adapters as commercial products.

BonitaSoft offers (in both the free and value-added versions of BOS) a library in excess of 100 pre-built adapters within the product plus more than 150 more on its open source community site, including those to either Talend or Mule ESBs.This allows BOS to support either true SOA orientation or “out-of-the-box” lighter weight integration such as data lookups and other directions designed/configured within Bonita Studio, but nonetheless working within the BOS environment.Access to additional adapters or new versions of existing adapters, however, is greatly enhanced by the bonitasoft.org community of over 10,000 developers.

The former is an example of the power of the community – having access to a vast and growing library of adapters. In contrast, consider the overall design and architecture of BOS (notably Bonita Studio and Bonita User XP.) Overtime the design of BOS (look-and-feel and supported functionality) will also be inevitably shaped and continuously enhanced by the user and developer community. Yet it requires stewardship from a product development and engineering team to set the direction for BOS as a commercially competitive product suite. This is where having the resources of a venture-backed, commercial software enterprise sets BonitaSoft apart from other OSS.

Ultimately, the combination of the two is what sets Bonita apart as a commercially competitive, enterprise-class complete BPMS, combined with an extended ecosystem of community-developed enhancements and value-added components.

Bonita Open Solution illustrates the benefits of a commercial open source model, through the combination of community-developed resources and commercially-competitive BPM capabilities developed internally by the BonitaSoft team.

Geek-Friendly, Business-Savvy

The challenge with business adoption of open source is eventually lack of business confidence – is it secure, accredited, stable? Again, open source should be inherently better, yet many firms still ban non-commercial open source for fear of security and stability issues. It is easy to get small components and class libraries in under the radar, but open source packaged application and full platforms (including BPMS) is still largely in the realm of early adopters – something exploited fully by commercial, closed-source vendors. BonitaSoft offers a true, freely-available open source option – one that has in the last year passed the One Million Downloads mark, with the vast majority of that occurring in the last two years (in terms of timing the original Bonita project launched in 2001, and BonitaSoft itself was incorporated in 2009.)

The challenge with business adoption of open source is eventually lack of business confidence – is it secure, accredited, stable? Again, open source should be inherently better, yet many firms still ban non-commercial open source for fear of security and stability issues. It is easy to get small components and class libraries in under the radar, but open source packaged application and full platforms (including BPMS) is still largely in the realm of early adopters – something exploited fully by commercial, closed-source vendors. BonitaSoft offers a true, freely-available open source option – one that has in the last year passed the One Million Downloads mark, with the vast majority of that occurring in the last two years (in terms of timing the original Bonita project launched in 2001, and BonitaSoft itself was incorporated in 2009.)

In addition, it offers a value-added service model in the form of subscription packs which provide for support as well as enhanced capabilities targeting larger scale production (performance monitoring, process templates, enhanced testing among other features.) This gets beyond the free-but-limited-functionality tease of commercial vendors, whether in the form a free modeler with no run-time environment, or a free download but restriction from production deployment. Your relationship with BonitaSoft need never involve the exchange of money. You can download a complete BPMS, have access to the full developer community, and deploy to production without ever paying them a dime.

Alternatively, your relationship with BonitaSoft can begin as a commercial transaction as with any BPMS market leader, complete with pre/post sales support, application hosting, training and enhanced capabilities – with the realizable benefits of these enhanced through the inherent extensibility of community-developed software.

The value of BPM is obviously beyond its own inherent functionality (i.e., specific application functionality built on the BPMS) but its ability to orchestrate processes across the enterprise, to serve as a transactional thread across multiple applications. For virtually all organizations, there is infrastructure and applications in place with which the BPMS needs the work. Yet invariably, there are also gaps. You may have SCM but not CRM; document management, but not the environment where you want to place yours next generations of content; SharePoint, but not the right portal platform where you’ll deploy your new web app.

This is where the “stack” vendors have sought to differentiate themselves, not necessarily as a better BPMS but as a bigger sum of parts. As a result, the mesh of interdependency has grown, with these platform vendors seeking greater ‘lock-in’ leverage, marketed as easier integration. The security model for the portal may be linked to the directory services used by CRM and the role resolution in the BPMS. If it all works as intended, then perhaps advantage can be found.

Yet there are two obvious but significant complications. First, most of these stacks were not developed in-house but through the assemblage/acquisition of various smaller vendors and 3rd party components, and consequently lack the level of pre-integration and common services expected. Second, however, is that despite aggressively discounted ‘bundle pricing’ the ultimate investment involves 5X or more what was anticipated for the BPMS alone.

The ability to plug-and-play, both/either commercial software and open source equivalents, new and existing without bias, offers many advantages over the current stack approach. The counter argument is that a best-of-breed assemblage of 3rd party components will lack the common services offered by those from a single platform vendor. Yet this argument falls short in the face of today’s less-than-integrated closed stacks, and misses the plug-and-play opportunities of common standards and OSS components.

BonitaSoft’s ecosystem of software partners and pre-integrated OSS components (see Figure 2 below) offers a compelling alternative to a single-vendor, vertically-integrated stack. To be clear, not all of these components come directly from BonitaSoft, but rather are available through their respective vendors and/or OSS hosting facilities such as SourceForge.net. The advantage offered by the BonitaSoft ecosystem is that integration has been pre-established and tested by developers participating in the bonitasoft.org community.

Figure 2: BonitaSoft Partner and OSS Ecosystem         

As discussed at the beginning of this paper, is the level of plug-and-play interoperability is hastened by the community model of open source development, yet is only possible today through the availability and leverage of standards. Standards help drive innovation by providing a demarcation lines around homogenous capability. Standards such as BPMN 2.0, XPDL and CMIS provide the critical leverage points for combining otherwise disparate OSS components into a comprehensive, full-function BPMS framework.        

Collectively, these components provide either OSS options for commercial best-of-breed options for any of the broadest closed stacks. This again extends and defines Bonita as not only a true BPMS in a stand-alone context, but as the BPM hub at the center of a complete SOA framework, when combined with open source (or commercially licensed) ESB platforms from either Talend (who also embeds Bonita in their own enterprise-class offering) or Mule. Continuing down the “stack” path, essentially any capability expected of a commercial software stack can be found in the Bonita ecosystem, including ECM capabilities from Alfresco, enterprise portal and “Enterprise 2.0” capabilities from Liferay, and so forth.

All of these will be supported by not only BonitaSoft as a commercial entity, but by the open community of users and developers. For this reason, the advantage is not simply the readily available, free components – which clearly does represent many advantages – but the fact that these environments are already widely supported in the enterprise and by a much more accessible talent pool (consultants, current and prospective employees) than is the case for proprietary, closed source stacks.

Bonita Open Solution supports a true model-driven metaphor for application development, enabling both the process itself and the run-time interface to be designed, deployed and executed within the same framework.          

Design-Connect-Run: Leveraging BPMN 2.0

The single-greatest core advantage of BPM today, and certainly the most engaging of business stakeholders, is the ability to build and deploy executable models. Increasingly, business executives generally unaware or otherwise unmoved by standards insist that BPM investments be based on BPMN. It has been far too easy for vendors to check this box, as BPMN remains a relatively loose standard (i.e., voluminous definition with limited conformance enforcement) and until recently lacked standard execution syntax or standard serialization other than XPDL.

In other words, every vendor has pursued their own interpretation of BPMN, and the only thing tying it together is XPDL. Yet the value there was nonetheless compelling – the ability to draw a model and then execute it; then as the business changes, only the model need, not all deployed code (along with required regression testing, etc.) This declarative ability remains the critical leverage point for BPM, but it is only as real as the standards behind. Propriety interpretation by vendors has in many (most) cases obviates any true round-trip ability.

Much to their credit, this is an area that BonitaSoft clearly takes seriously. In this latest release, this includes model interchange both within the Bonita environment (i.e., round-tripping) but also between environments compliant in XDPL or BPMN. This is a significant market advance, and while ironically this area of standard adoption has been faster for commercial vendors than the open source community, both have left much to be desired.

Here Bonita stands out visibly. As my colleague Bruce Silver describes, Bonita is “so far the most successful in actually implementing BPMN 2.0 based model interchange. ” This type of praise does not come easy. While Bruce may not go out of his way to criticize other vendors, there had been few chances to praise those who truly leverage the promise of BPMN 2.0 – the silence had been deafening.

This stand-out quality is not a matter of “pretty good for open source” but clearly places Bonita high among BPMS market leaders. It enables a “design-connect-deploy” model (which should be) expected of any best-of-breed BPMS – where all logic is configured in the designer; connections, screens and forms are built in single unified environment; and where deployment is essentially one- click rather than an aggravating series of ANT scripts.

Open Source 3.0 is indeed very likely here today (finally.) It is not simply the next generation of open source, it is leveraging community-driven benefits with commercially led software leadership. As a venture-backed concern with over $17 million invested in the last two years, we expect BonitaSoft to be a visible innovator and powerful disruptor for 2012 and beyond.

For more information about BonitaSoft visit http://www.bonitasoft.com

BPM.com has been the Internet’s top destination for articles, news, research and white papers on Business Process Management and workflow. BPM is a process-centric approach to managing business operations, evolving from earlier and existing technologies and practices including business process re-engineering, workflow management, and document management.

        

About the Author

Nathaniel Palmer is Editor in Chief of BPM.com, as well as CTO of Business Process Management, Inc. and the Executive Director of the Workflow Management Coalition. Previously he was Director, Business Consulting for Perot Systems Corp, and prior to that spent over a decade with Delphi Group as Vice President and CTO. He is author of over a dozen books on BPM and IT led transformation, including The X-Economy: Profiting from Instant Commerce, Social BPM, The BPM and Workflow Handbook, and The Encyclopedia of Database Systems among others. He has been featured in numerous media ranging from Fortune to The New York Times, plus over 100 by-lined articles in IT publications such as CIO and InformationWeek. He has also been featured as a guest expert on National Public Radio and World Business Review.

He can be reached at This email address is being protected from spambots. You need JavaScript enabled to view it.

 

 

  • No comments found

Leave your comments

Post comment as a guest

0
terms and condition.